Environmentally Influenced Urbanisation: Footprints Bound for Town?

Abstract

Over the past 30 years, urbanisation has been a prominent phenomenon and various drivers have been proposed to explain it. Very few have suggested that the degradation of the rural environment was one of them. This paper explores the human–environment interface by focusing on the portrayal of these concepts within scholarly literature. A systematic literature review was conducted and 147 articles were examined to determine the direction of the link between the environment and human mobility, and if urbanisation was featured. The results demonstrate that equal attention is paid to both directions of the environment–mobility link. Of the articles reviewed, 40 per cent focus on urbanisation, but 93 per cent of those portray urbanisation as a forcing on the environment, rather than an impact of environmental degradation. The lack of support for environmentally influenced urbanisation can be explained by coupled system complexity, disciplinary research and the silence of those most likely to endure environmental change. Understanding these relationships is paramount to the promotion of adaptation without eroding resilience or further degrading environments.

1. Introduction

Since 2008, for the first time in the history of planet Earth, the number of people living in urban areas exceeds those residing in rural areas (United Nations, 2008). At the same time, physical environments and ecosystem services are increasingly portrayed as degraded, with the potential to push populations away from homelands that can no longer sustain livelihoods. Very few scholarly articles have drawn a direct link between the phenomena of urbanisation and environmentally induced mobility. This review explores these concepts side-by-side, with the aim of evaluating the degree to which they are linked. The concept of environmentally influenced urbanisation (EIU) is set in the context of a broader discussion on environmentally influenced mobility (EIM).

The terms EIU and EIM are both presented in this section. Also featured here is the concept of a ‘footprint’ that shows how human mobility and urbanisation are influenced by changes in the physical environment. Section 2 describes a wide range of theoretical foundations that have implications for the environment–mobility nexus and urbanisation. Sections 3 and 4 describe the methodology and results focusing on the direction of the environment–mobility link and how urbanisation is addressed. Section 5 concludes on how closely environmentally induced mobility and urbanisation are related and highlights implications for adaptation and policy in a sustainable future.

1.1 Urbanisation

As of 2010, according to the United Nations, there were 6.9 billion people on Earth. Of these, 3.5 billion, 51 per cent of the total, live in urban centres. The 50 per cent threshold was reportedly crossed in 2008 and few scholarly documents hail this major change in global settlement patterns. Urban population, as a percentage of the total, has risen steadily since 1950, when it was under 30 per cent. Projections suggest that it will go beyond 70 per cent by 2050. Since the 1970s, the population data most readily amenable to intercountry comparisons have come from the United Nations Population Division. Despite the challenges linked to definitions and geographical boundaries, the trends indicated by these data are generally reliable (Montgomery, 2008).

Urban issues have rapidly gained salience in scientific research. There are at least 54 active electronic journals with ‘urban’ or ‘city’ in the title. The oldest of those still active include Urban Studies (archives dating back to 1963), Urban Education (1965), Urban Review (1967), Education and Urban Society (1968) and Urban Lawyer (1969). Few of these journals have a direct focus on the physical or natural environment. Only seven, all created within the past 25 years, mention some aspect of the environment in their titles.

A bibliographic search featuring the term ‘urbanisation’ or ‘urban growth’ in the title (excluding less pertinent research areas such as public health, engineering or architecture) retrieved 1960 peer-reviewed articles published between 1907 and 2010. The largest proportion originated in physical or environmental sciences (46 per cent). The most prolific institutions were the Chinese Academy of Science (China is home to the greatest volume of urban dwellers with one of the fastest urban growth rates) and Arizona State University (located in one of the fastest-growing states in the US).

A high proportion of articles on urbanisation contributed by physical or environmental science might lead one to conclude that environmentally induced migration to urban centres was a frequent topic of exploration. Few articles, however, focused on this; the vast majority explores deleterious forcings of urbanisation on ecosystems and services.

1.2 Environmentally Influenced Mobility (EIM)

Individuals have sought refuge from extreme environmental events for millennia. Current estimates of environmentally influenced mobility abound, based largely on back-of-the-envelope calculations and in-depth case studies in isolated areas. The IFRC estimates that there are 25 million ‘climate migrants’ (International Federation of the Red Cross, 2001). While sceptics claim that there are no ‘climate migrants’ because other drivers, such as economics, dominate the push factor (Black, 2001), ecologist Norman Myers proposes a ‘conservative’ estimate of 150–200 million over the next 50 years, attributed primarily to global climate change (Myers, 2002). Christian Aid has recently predicted 1 billion environmentally driven migrants by 2050 (McNamara, 2007).

From 1958 to 2009, at least 500 authors wrote hundreds of documents to propose specific causes explaining the links between human mobility and the environment (Erway Morinière, 2009). Debate is raging on the forces driving this mobility. Just as economists have long stressed purely monetary influences at both origin (poverty or unemployment) and destination (job opportunities and profit maximisation) (Mougeot, 1992), each discipline favours a different driver. Natural resource managers consider environmental degradation to be the major trigger (Doos, 1994). Disaster risk scientists contend that natural hazards are among the most forceful drivers (UNOCHA and IDMC, 2009). Climatologists explain the phenomenon in terms of changes in temperature (IPCC, 2007). Development workers point to mega projects that displace entire communities (Saxena, 2008). Urban planners have not proposed an explanation.

Most researchers concede that, whatever the specific driver or status, the probability is high that human mobility will accelerate. To date, however, no recognised dataset exists to monitor this phenomenon.

Environmental degradation does not readily appear in longitudinal global data. Commonly referring to land, degradation can take many forms (water, wind, chemical and physical) and has many causative factors such as deforestation, removal of natural vegetation due to overgrazing, agricultural activities, overexploitation and bio-industrial activities (Bridges and Oldeman, 1999). The only systematic estimates of global degradation are from the Global Assessment of Human-induced Soil Degradation (GLASOD) in 1991 and the Global Assessment of Land Degradation and Improvement (GLADA) in 2008. The former found 15 per cent of the earth’s land surface degraded and the latter found 24 per cent of it degrading between 1981 and 2003, although not with the same geographical distribution (Bai et al., 2008). Despite different methods, there is little dispute regarding the general trajectory of environmental decline that is described.

1.3 Defining the Footprint

W. Rees (1992, p. 121) made popular the concept of an ‘ecological footprint’ to signify the “total area of land required to sustain an urban region.” Rees suggested that there may be inadequate natural capital and carrying capacity to sustain current ecological footprints, especially under anticipated rates of urbanisation. Hundreds of scientists not inventoried here have carried that footprint inquiry forward to account for the historical influence humans have had on nature—at least once in regard to migration (Cairns, 2004).

No use of the term ‘footprint’ was found in scholarly literature to reflect the physical trace left by humans in motion—regardless of the driver of their mobility. In this study, ‘footprint’ constitutes any link between an aspect of the physical environment and a form of human mobility. When the link is found to start in the environment—even if it is only one of a complex set of influences or drivers—the resulting footprint confirms the forcing. Tracing the footprint will cast greater insight on the interface between humans and the environment.

Natural ecosystems—particularly when productive capacity is inappropriately used—may exert a push on humans, driving them away from their homelands when that capacity becomes insufficient (Rees, 1996). Such pushes can arise suddenly, as in the case of geological events (such as, earthquakes and volcanic eruptions) or meteorological/hydrological events (such as storms and floods). They can also arise slowly, as in the case of the long-term processes of degradation, desertification or sea-level rise. Another important push is scarcity of natural resources, such as fresh water or arable land. Development and conservation projects often drive populations from their homes near waterways (i.e. dams) or within newly protected areas. These are very different pathways with a similar result: an environmentally influenced footprint.

Human mobility within this research is entirely geographical; it does not entail any vertical movement between socioeconomic classes. Mobility encompasses purely voluntary migration, forced displacement and every step of the continuum between these extremes. Human mobility has many forms, including but not limited to migration stocks, refugee flows and urbanisation. Mobility can entail successful adaptation, the failure to adapt, a temporary coping strategy or last-minute flight. Most often, households with the scantiest resources (for example, socioeconomic capital) are left behind in degraded environments, unable even to consider mobility. In this respect, the most vulnerable are excluded from the present discussion.

2. Theoretical Foundations

The world’s most pressing problems with long-term wide-area impact, such as ‘forced migration, poverty and environmental crises’, are too complex to be addressed by single disciplines (Max-Neef, 2005). Over the last decade of the 20th century, the growing interdisciplinary movement in scholarly research has sought a holistic understanding of phenomena for a common purpose, such as policy. A review of multidisciplinary efforts, however, suggests that these efforts remain “limited in impact on theory building for coping with the changing human condition” (Rosenfield, 1992, p. 1343). A subsequent transdisciplinary movement insists on the fusion, rather than the mixing, of disciplines and theories with a result that exceeds the simple sum of each (Lawrence and Després, 2004). Transdiciplinarity lends itself most readily to the exploration of complex problems.

Aiming for transdisciplinary insight, the present analysis draws on several theories. Understanding the tenets of many pertinent theories permits a more complete understanding of environmentally influenced mobility (EIM) or urbanisation (EIU). Starting with the most holistic, these theories are drawn from human and political ecology, climate change, anthropology, demography, migration and urbanisation. The theory sets are summarised in Table 1, in light of environmental influences on human mobility and urbanisation. The intent is not to provide a thorough review of an exhaustive set of theories but to set side-by-side elements of a few that have clear implications for the concepts under study. Migration and urbanisation theory, the main features of this paper, are presented more thoroughly in the next sub-sections.

Table 1.

Theoretical implications for environmental influenced mobility (EIM) and/or urbanisation (EIU)

Theory sets	Implications for EIM/EIU
Ecology: political, human	Systematic inclusion of environment, population (culture and character), organisation and technology as major drivers of mobility Policies on land tenure, agriculture and migration are important factors of any analysis of human mobility
Climate change	High risk associated with urban growth, both to ecosystems and to inhabitants Ongoing debate: influence of urbanisation on the environment (for example, urban heat island effect)
Anthropology, cultural theory	Social bonding and need for order direct individual membership to one of four groups: individualists, hierarchists, egalitarians and fatalists, each producing a unique mobility profile
Demography	Malthus vs Boserup on implications of resource scarcity for source regions of mobility and/or urbanisation
Migration	Push and pull theory insists on an impetus at both source and destination, but the economic pull of a city is routinely given greater weight than an environmental push Economists focus on money as an explanation for mobility to the exclusion of other factors, but there is growing agreement that rapid population growth and low economic growth do not automatically induce migration
Urbanisation	Differences in expected income drive rural-to-urban migrationPast few decades, urban investment in industry and services drives migration Ecological impact of urban areas portrayed as major driver of climate change

2.1 Migration

The major challenges to migration theory are that: migration is the only demographic variable heavily influenced by regulation; and, it is one of the least studied and standardised of all demographic variables. In fact, migration data are rarely collected solely, or even primarily, for demographic purposes and had, until recently, been relatively neglected by traditional demographers (Poulain and Perrin, 2009).

During the 1880s, when migration was still distinguished by ‘lawlessness’, Ravenstein (1885) documented ‘Seven laws of migration’, based on research compiled from 20 countries. Five of these relate to the issues under study. Most migration entails a short distance; those going far are attracted to centres of commerce and industry. This has obvious implications for internal and South–South migration as well as urbanisation. Migration occurs in waves (rural to suburban, suburban to city centres). Each main stream of migration produces a counter-stream (for example, unemployment in urban centres due to high demand from migrants triggers out-migration feedbacks). Rural inhabitants migrate more than urban inhabitants. Positive and strong relationships exist between technology (industrialisation and urbanisation) and migration. Drivers include oppressive policy, taxation and unattractive climate.

As one of the main attributes of his work, Lee (1966) enriched Ravenstein’s laws by documenting the ‘push and pull’ theory, whereby at least two sets of drivers or enabling factors always interact: one at the source, pushing the potential migrant, and another pulling the migrant towards the eventual destination. While economists insist that money can explain all human mobility, contemporary migration theorists generally agree that it is impossible if not disingenuous to isolate drivers of migration (Lonergan, 1998; Black, 2001).

Lee further suggests that migration is most efficient when drivers at the origin are negative, such as in environmental extremes. Bilsborrow (1992) developed a decision flow model exploring household strategies that include off-farm employment, seasonal migration to urban areas and finally, permanent migration. Out of 10 household coping strategies among Sudanese listed by Watts and Bohle (1993), three involve mobility: sale of labour power, dry season farming and permanent migration. Cutler (1986) describes adaptive strategies in Ethiopia that include labour migration with mass migration as a last resort. De Waal and el Amin (1986) describe three sets of strategies of households in Darfur; the first two contain various types of migration and the final stage is destitution. Corbett (1988) also places distress migration within a final stage of destitution.

2.2 Urbanisation

The technical dictionary Demopaedia (United Nations, 1982) describes urbanisation as “growth in the proportion of persons living in urban areas”. The single most distinguishing and measurable characteristic of a city is size; cities can grow in at least three different ways

—they can expand in area (for example, urban sprawl inevitably denoting some loss of natural habitat);

—they can increase in total population (number of inhabitants within the official boundary); and

—they can rise in importance—economic, cultural or other.

Authors have portrayed urbanisation through history in many ways. Some refer to an unrelenting switch of labour from agriculture to industry (Bairoch, 1985). Urbanisation and industrialisation are often used as synonyms (Todaro, 1997). A city’s total population can also grow in three ways

—by the natural increase of current inhabitants of a city (i.e. from the inside). The UN’s Department of Economic and Social Affairs (UNDESA) estimates that natural increase explains 60 per cent of urbanisation in developing countries, especially in Latin America and Asia (UNDESA, 1985). Evidence nonetheless attests to lower total fertility rates in urban than in rural areas of developing nations (Bongaarts et al., 1984; Lee and Farber, 1984).

—by the (in-)migration of new inhabitants from the outside. Referred to by some as the greatest mass movement of human history, migration to cities is considered a primary contributor to urbanisation (Bairoch, 1985; Todaro, 1997). Quantitative analyses on the subject are dated; estimates attribute from 21 per cent to 75 per cent of urban population growth per country to migration,¹ the higher percentages reserved for least-developed nations (Newland, 1980; Chen et al., 1998); and

—by spatial expansion, or the annexation of surrounding settlements (Kasarda and Cremshaw, 1991).

In this review, urbanisation refers to the rate of urban growth: the number of inhabitants of bounded urban areas (Kasarda and Cremshaw, 1991) arriving from outside.

Urbanisation, one of civilisation’s ‘finest achievements’ (Rees, 2005), is considered primarily to exert an economic pull factor on migration, rather than result from a rural environmental push. This appears to neglect a more thorough exploration of what factors have driven, or may continue to drive, rural populations into cities.

Push factors for rural-to-urban migration are rarely discussed in contemporary academic literature. Conventional drivers of urbanisation include the rise of agriculture, trade and political organisation. Newland (1980) presents three main drivers of urbanisation circa 1980: unabated population growth, resource scarcity and technology. Montgomery (2008) cites four main influences: globalisation, decentralisation, poverty alleviation and climate change. According to the Todaro migration model, differences in expected incomes, based on perceptions of urban wages, are the main factor determining rural-to-urban migration (Todaro, 1997). Some researchers blame high investment in urban industry and services for the urban growth of recent decades (for example, Satterthwaite, 2009). For dependency theorists, capitalism is a unique social system and the allure of city wealth will naturally catalyse social inequality and rural-to-urban migration (for example, Timberlake, 1987).

Other urban theory perspectives suggest that cities benefit the environment because they concentrate half the world’s population and impact on small areas (Crane and Kinzig, 2005). Urbanisation has benefited technological innovation while heightening economic inequality and concentrating hazard risk. The failure of urban centres to meet migrant expectations, often referred to as an ‘urban shadow’, can catalyse ex-urban mobility (i.e. movement beyond the suburbs) thereby serving as an incubator for disasters and casting a negative feedback on rural areas (Mulvihill and Ali, 2007).

3. Methods

With the overwhelming volume and speed of digital information generated by the current generation of scientists, synthesis is paramount. In 2009, a multidisciplinary peer-reviewed journal entitled Research Synthesis Methods was created to document techniques that synthesise scientific information. Research synthesis, as a science, emerged only in the 20th century and comprises both systematic reviews and meta analyses guided by the growing appetite of decision-makers and practitioners (Chalmers et al., 2002). Glass, who coined the term ‘meta analysis’ (MA), insisted that, although both are vital components of science, synthesising knowledge to discern the broader outlines is a role more important in this era than adding bits of facts (Glass, 1976). Capitalising on what is already known is considered not only an academic responsibility but also proof of scientific self-discipline (Chalmers et al., 2002).

Drawing on the field of information technology (IT), systematic literature reviews (SLRs) are the method of choice for compiling evidence (Kitchenham et al., 2009). Going beyond the informal ad hoc literature review conducted as a background to most research efforts, a SLR applies rigorous methodology to capitalise on a large body of documented empirical evidence; it is a prerequisite for meta analysis. Procedures for conducting a SLR include: problem statement, review protocol, search for evidence, selection of studies, quality assessment, data extraction, synthesis and reporting (Kitchenham, 2004). All SLRs enter criteria about selected studies into a database for analysis.

The present SLR was conducted using the Web of Knowledge (WoK) portal. ² The first step was to run searches for documents systematically linking some aspect of the physical environment to a manifestation of human mobility. To constrain the review, the initial searches were limited to journal articles with some aspect of each element specifically named in the title (rather than topic) in order to identify articles in which the exploration of these issues was explicit. For ‘environment’, the filter³ was set up to capture any aspect of its physical or natural elements, including specific hazards or disasters, climate change, slow onset processes and conservation or dam projects. For ‘mobility’, the search string⁴ included use of any form of the following terms: mobility, migration, refugee or displacement. The two searches were combined (excluding non-human migration), resulting in a preliminary set of 643 articles.

Each article was filtered to include only those demonstrating dynamic elements: changing environments or moving or growing populations. This not only further confirmed the presence of the environment–mobility link, but also excluded articles in which the link was descriptive rather than directional. This process produced a set of 71 articles which were entered into a database. Finally, all other articles that cited (one generation forward) each of the 71 were explored to capture any pertinent work that might have been omitted by the initial WoK (title) search. An additional 511 articles were reviewed and 76 more were added to the database. Articles that were retained feature the environment–mobility nexus as a main argument, not just a peripheral association mentioned in passing or as one of many results.

The final set of articles in this analysis includes 147 spanning nearly 75 years (1936–2010). The number of articles meeting the criteria stayed low through the 1980s (3, 2 and 6 for prior to 1970s, 1970s and 1980s respectively). The 1990s produced 26 articles and the first 11 years of the 21st century, 105.

4. Results

The initial analysis consisted of a review of the 147 articles to determine the direction of the link between environment and mobility, the precise elements studied, the inclusion of urbanisation and the geographical focus of the research. Each of these elements is next discussed in turn.

4.1 Direction of Link

The direction of the environment–mobility link (see Figure 1) can portray either: human mobility as a driver of environmental degradation (type A); or environmental degradation as a driver of human mobility (type B).

Figure 1.

Types A and B of the environment–mobility link.

Among the 147 articles reviewed, the two directions received the same level of attention: 48 per cent for type A, human mobility degrading the environment (n = 70) and 52 per cent for type B, degrading environments triggering human mobility (n = 77). When analysed through time, however, the characterisation of this direction is much less evenly distributed. Prior to the 1970s, only three articles were found; all of them portrayed degrading environments triggering human mobility (type B). The 1970s produced only two articles, one of each direction. The 1980s produced six qualifying articles, but the ratio of type A to type B links was two to one. In the 1990s (26 articles), the trend switched with 63 per cent showing type B links and 38 per cent type A. Finally, during the period 2000–09 (105 articles), the proportions evened out to exactly 50 per cent each. These shifts are likely to reflect the evolution of knowledge as well as trends in funded research.

4.2 Geographical and Temporal Foci

The 147 articles have diverse geographical foci. Thirty-six per cent (n = 45) feature an environment–mobility link in the Americas, followed by 30 per cent in Asia and 20 per cent in Africa. A small fraction features the Middle East, with Europe in last place.

Six of the articles are Paleolithic studies on mobility motivated by large-area climatic forces. Late Holocene populations, disrupted by extreme climatic events, were thought to have been able to shift to lower subsistence levels, including departure from urban centres (DeMenocal, 2001). It is likely that early Holocene hunter-gatherers were forced to migrate from Spain due to severe drought conditions (Gonzalez-Samperiz et al., 2009). Abrupt loss of land from the Laurentide ice-sheet collapse and subsequent sea-level rise purportedly triggered the Neolithic expansion of peoples across Europe (Turney and Brown 2007). Neolithic settlement patterns are also closely associated with rising sea-levels in eastern China (Chen et al., 2008) and farmers’ movement from the Black Sea to Europe is linked to an epic flood event (Giosan et al., 2009).

In the US, the earliest contemporary articles positing links between the environment and human mobility describe internal and cross-border movement driven by drought (Rowell, 1936; Taylor and Vasey, 1936). These documents contain the earliest recorded use of the term ‘drought refugee’. McLeman (2006) adds detail and insight to these records, describing the 1930s dust bowl and out-migration from Oklahoma.

4.3 Elements of Environment and Mobility

Environment

Fourteen environmental factors are featured in these 147 articles. Up to three were explored in each article; the most commonly studied overall, in order of frequency, include climate change, degradation, drought, flooding, deforestation and resource scarcity. Prior to the 1970s, the selected articles exclusively portrayed drought as a driver. In the 1970s and 1980s, earthquakes, flood and degradation were added to the mix, along with pollution and climate change. The 1990s ushered in slow-moving drivers: resource scarcity, deforestation, sea-level rise, desertification and conservation projects. No new environmental drivers were added during the period 2000–09.

Type A articles (mobility causing environmental degradation) are dominated by explorations of climate change (35 per cent), followed by deforestation (16 per cent), degradation (14 per cent) and flooding (12 per cent). Type B articles (environmental degradation causing mobility) explore a mix of environmental factors focusing above all on drought (19 per cent), climate change, degradation (12 per cent each) and resource scarcity (11 per cent).

Mobility

Mobility has many facets; these articles portray mobility in at least six different ways. In order of frequency, they include urbanisation, out-migration (mainly departing rural areas), general immigration, displacement (or homelessness), refugee and resettlement. Urbanisation has always been the most commonly studied of these manifestations, but during the period 2000–09 it was studied nearly twice as often as any other manifestation. Refugees have not been a major topic of concern in this context because, by a UN resolution of 1951, it is a phenomenon explicitly tied to a “well-founded fear” of political—not environmental—“persecution.”

The vast majority of type A articles (mobility triggering environmental degradation) relate to the specific exploration of urbanisation (80 per cent) and, to a lesser extent, immigration (10 per cent). Type B articles cover a wider range of foci; the majority explore rural-to-urban/out-migration (34 per cent) and displacement (20 per cent).

4.4 Inclusion of Urbanisation

A total of 58 (40 per cent) of the 147 articles explicitly address urban population growth. Depending on the direction of the link they support, I refer to these henceforth as type A:urban or type B:urban. The vast majority of the 58 portray urbanisation as a type A phenomenon exerting a negative influence on physical ecosystems (between 50 and 100 per cent of the articles published during each period). The number of type A:urban articles grew from nil prior to the 1970s, to 50 per cent in the 1970s to 98 per cent in 2000–09.

Type A:urban articles explore climate change more frequently than any other driver (39 per cent). The second most common environmental aspect studied is flooding (14 per cent). Consensus among the type A:urban articles holds that the major human-induced urban influences on climate include land use change (from rural to built environments) and greenhouse gas emissions (Martine, 2009). There is nonetheless significant debate on the exact contribution of cities to emissions (Grimm et al., 2008) and the precise contribution of the poorest cities (such as in developing countries) and slum dwellers (Martine, 2009; Satterthwaite, 2009).

Only four of the original 58 articles (7 per cent) linked urbanisation to environmental drivers of mobility (Osterling, 1979; Main, 1990; Doos, 1994; Berhanu and White, 2000). The four type B:urban articles are discussed in greater detail in section 5. The scant literature on urbanisation—as driven by or linked to environmental push factors—was more common in the 1970s than in the other decades.

5. Discussion

For many decades, scientists who study these topics appear to have accepted the inevitability of unrelenting urbanisation with its mixture of opportunities and perils. As a result, they have focused almost exclusively on ecosystem impacts or on challenges in improving urban livelihoods, resilience and productivity. According to some 20th century scholars, however, urbanisation is not inevitable. Afshar (1998), citing examples from Thailand and Pakistan, argues for counter-urbanisation: the ruralisation of cities and rurbanisation, a hybrid of rural and urban. Others claim that urban planning and policy have succeeded in slowing or reversing urbanisation trends in many settings (for example, Israel and eastern Europe) (Brutzkus, 1975). Some suggest that Third World urbanisation results from inappropriate—therefore evitable—development models (Rees, 1992). The concept of sustainable cities implies that they may not be, thus an oxymoron (Rees, 1997). Before fully examining urbanisation’s roots, academic literature has already moved on to explore reverse dynamics: the ex-urban phenomenon (Mulvihill and Ali, 2007; Robinson and Brown, 2009) and urban shrinking (Kabisch et al., 2006; Haase, 2008; Pallagst et al., 2009; Rieniets, 2009).

Scholarly papers addressing urbanisation influenced by the environment are scant; only five such articles appeared in the compilation. The first article discusses a major earthquake that struck Peru in 1970, after which, survivors started moving towards Lima (Osterling, 1979). The author links the disaster to out-migration, ‘modernisation’ and urban acculturation. In the second, dam development completed in 1976 in northern Nigeria was tied to population growth in the state capital, Kano (Main, 1990). Displaced households lost agricultural land within the flood zone of a planned dam. Up to 75 per cent of those who had moved to Kano did so because of the dam.

According to the third article, environmental degradation is an escalating driver of global food insecurity; urbanisation (resulting in loss of arable land) is a major negative forcing (Doos, 1994). Growing populations require urban/industrial development that consumes half the current agricultural land. Doos proposed a feedback loop whereby food shortages causing environmental degradation triggered urbanisation, which in turn caused further degradation of the immediate environment. Based on insufficient purchasing power, Doos claimed that the likelihood of major migration is strong; adding water scarcity increases the risk, particularly of South-to-North migration.

In a fourth article, famine-linked female mobility to urban centres was explored in Ethiopia for the period 1960–89 (Berhanu and White, 2000). Sharp increases in rural-to-urban migration of women to Addis Ababa were associated with both drought and civil strife.

In the final fifth article, a clear link is drawn between Bangladeshi floods and squatters in Dhaka (Lein, 2000). A literature review on the environment–mobility interface that studied 321 articles since 1958 by over 87 academic journals featuring 506 authors (Erway Morinière, 2009) produced the final type B:urban document.

These studies of environmentally influenced urbanisation (EIM) are noteworthy for their singularity in the scholarly literature. They are sufficiently separated in time (1979, 1990, 1994 and 2000), in space (Peru, Nigeria, general/global, Ethiopia and Bangladesh) and by driver (earthquake, dam construction, environmental degradation, famine and floods) to make weaving any further patterns between them impossible. Their only commonality is a type B link between the environment and urbanisation.

This is disquieting: how does such an obvious association escape scholarly attention? Since type A and B links are evenly distributed when exploring EIM in general, why are scholarly articles exploring type B:urban (EIU) links so few? Why does the type A link gain more attention in urban literature? A few explanations are next proposed, each followed by a short discussion of their plausibility.

First, there may have been an insufficient search for evidence. A common problem of systematic or other literature reviews is that the evidence may not be fully representative of all studies conducted. It is often called the ‘file drawer problem’ in which sit 95 per cent of all studies conducted—those with non-significant results (Arnqvist and Wooster, 1995). Although the Web of Knowledge is a reputable search engine for peer-reviewed literature and the searches applied were thorough and well documented as a systematic literature review (SLR), some articles may have been inadvertently excluded (such as those in languages other than English). Further biases reinforced by independent analysis are possible during the coding of articles. Any environmental stress added to the search strings will inevitably produce more, or a different set of, articles. This argument cannot be entirely ruled out.

Secondly, there may be pertinent articles in grey literature. The environment–mobility nexus may be well explored, but by grey or practitioner literature—with scientific literature lagging behind. Publication bias—when significant positive results are given precedence in scholarly literature—can be controlled in part by a parallel review of grey literature and conference proceedings (Kitchenham, 2004). For example, a few unpublished reports highlight the type B:urban link. Wood (2001) states that “the most significant ecomigration in the Third World is that to and from Third World cities” and Ezra and Kiros (2001) join Berhanu and White (2000) to explore rural environmental pushes as being more important than the pulls of Ethiopian cities.

Further, the Forced Migration Review, recently featured a special issue in 2010 entitled ‘Adapting to Urban Displacement’. At least four authors make clear reference to the type B:urban footprint. Albuja and Ceballos (2010) discuss land scarcity as driving urban growth in Colombia. Tibaijuka (2010) portrays disasters as triggers of rural-to-urban migration. Zetter and Deikun (2010) state that the number of people arriving in cities after escaping environmental disasters is increasing. They describe a United Nations High Commission for Refugees (UNHCR) policy document that protects refugees in urban settings which reports that urban ‘internally displaced persons’ (IDPs) make up nearly half of all IDPs (over four million) and are driven by environmental disaster.

In summary, grey literature provides vivid type B links and rich case studies. Although a quantitative estimation of their prevalence is not possible, continued theoretical support from scholarly literature to the practitioner debate, as already described, may be useful.

Thirdly, the complexity of the subject matter transcends disciplinary perspectives. Urbanisation, human mobility and environmental degradation are all, individually, extremely complex phenomena. Disciplinary approaches to problem solving are still the norm in higher education (Max-Neef, 2005). In fact, the ‘higher’ one gets—the more specialisation achieved—the more scholars function in isolation. Institutional obstacles are among the most persistent weaknesses that exacerbate knowledge and theory production. The study of complex systems in a transdiciplinary manner is also seen as unrewarded risk-taking by many academicians (Rosenfield, 1992). Disciplinary science to date has not taken the risk to explore fully this interface.

Perhaps, too, disciplinary science can be partially blamed for the negative view of the city. The negative press that cities get from climate change and environmental scientists (see Table 1) may not provide a compelling incentive to trace urban growth to environmental degradation.

The complexity of environmentally influenced mobility or urbanisation is such that it may be impossible to isolate drivers (Lonergan, 1998); perhaps the lack of type B:urban studies confirms this. On the other hand, it may be that every reference to labour or other migration to cities has elements of the environment that were simply not untangled: “all migrants are environmental migrants in some form or another” (Carr, 2005, p. 944).

It is also possible that contemporary science is so engrossed with the quest for creative and sustainable solutions in a predominantly urban world that exploring why urbanisation is relentless is seen as a task with little added value in the greater scheme of things. This is an entirely human but dangerous and inefficient position to hold. A more thorough and regularly refreshed understanding of the exact drivers of urbanisation provides solutions in and of itself. For those who deem urban growth to be one of the 21st century’s major concerns, a clear grasp of its exogenous drivers offers the only hope of eventually curbing the phenomenon. For those who see urbanisation as a necessary evil or utopia, this understanding will help to channel human flows towards more healthy urban ends.

A final explanation of this gap is that it may be hiding something more insightful. As a famous demographer eloquently stated

Critically interpretive research begins with a series of negative questions: what is being hidden from view in the official [migration] statistics? Whose economic or political interests are reflected in the kinds of records kept? What is thought hardly worth counting at all? What can this tell us about the collective invisibility of certain groups and classes of people? (Scheper-Hughes, 1997, p. 220).

Many of those currently residing in cities were likely to have been driven from their rural homes for a variety of reasons that include environmental degradation. Their footprints bound for town are defended at the very least by the five studies resulting from this SLR. Populations confronting environmental stress that pushes them to town may have the frailest voices, making them—to some—not worth counting.

6. Conclusions

This review has compiled a database of 147 peer-reviewed articles in an attempt to characterise an active link between the physical environment and human mobility. Each article was studied to capture the direction of the link and the inclusion of urbanisation. The following conclusions result from this review.

Large bodies of scholarly literature explicitly, but separately, address urbanisation and environmental degradation. Peer-reviewed literature documenting a clear link between the two is scant. The link between environments and mobility is bi-directional: type A links occur when human mobility contributes to the degradation of the environment; type B links happen when degrading environments edge humans into movement.

Although type A and type B links are overall given the same level of attention in scholarly literature, degrading environments as contributing to urban growth (type B:urban link) are not readily supported. Only five articles were found to link explicitly environmental drivers to urban growth. Beyond biases in the review, this gap runs contrary to many of the theories explored in section 1. Human ecology and demographers encourage a view of flight as a standard evolutionary response to scarcity and environmental degradation; contemporary climate and environment scientists confirm sufficient scarcity and degradation to trigger this flight; urbanisation has slowed only in the wealthiest of nations—and yet, the footprints bound for town are hard to trace. The fusion of the latter theories with complexity theory and political ecology readily offer support. Through complexity theory, science will systematically explore non-linear drivers and track the influence of the environment on every economic decision to migrate. Political ecology will seek the broader often socially constructed causes of urbanisation, exploring power struggles and exposing inequalities for the push and the pull.

Degrading physical environments appear to have influenced and may continue to influence human mobility and urbanisation. The most concrete examples include sudden-onset natural hazards that make homelands uninhabitable (i.e. volcanoes and earthquakes) but also slow-onset phenomena such as sea-level rise. Without clearly compiled evidence at the global scale, this remains, to date, conjecture. There are likely to be deep footprints bound for towns across the globe, but details on their size, shape, depth and orientation require much further untangling from the complex set of influences that have set them in motion.

Appropriate policies are nonetheless required to protect resource-dependent rural livelihoods from reaching the point of no return, thereby pushing households towards town. Such policies are inevitably complex and replete with economic and social constraints. They could include practices that span the protection of agricultural lands from sustained chemical abuse, to water rights, land tenure and culturally appropriate resettlement. It is also important that the academic community rise to the analytical challenges of global-scale transdisciplinary research, while the United Nations lobbies more proactively to build or reinforce official record-keeping of internal mobility as critical components of sustainable solutions and the protection of the right to remain securely in one’s homeland.

Footnotes

Acknowledgements

This study is based on a series of related footprint research efforts supported by the Stockholm Environment Institute (SEI, Oxford), the Office of Arid Land Studies (OALS/UA), the Marshall Foundation (UA), the Public Entity Risk Institute (PERI) and others. The usual disclaimers apply. The author is grateful for comments from D. Plane, K. Waser and anonymous referees on a previous version of this paper.

Notes

References

Afshar

(1998) Balancing global city with global village, Habitat International, 22(4), pp. 375–387.

Albuja

Ceballos

(2010) Urban displacement and migration in Colombia, Forced Migration Review, 34(February), p. 10–11.

Arnqvist

Wooster

(1995) Meta-analysis: synthesizing research findings in ecology and evolution, Trends in Ecology & Evolution, 10(6), pp. 236–240.

Bai

Dent

Olsson

Shaepman

(2008) Global assessment of land degradation and improvement. Report No. 2008/01, I. W. S. Information, Wageningen.

Bairoch

(1985) De Jericho a Mexico: villes et economie dans l’histoire. Paris: Gallimard.

Berhanu

White

(2000) War, famine, and female migration in Ethiopia, 1960–1989, Economic Development and Cultural Change, 49(1), pp. 91–113.

Bilsborrow

R. E.

(1992) Population growth, internal migration and environmental degradation in rural areas of developing countries, European Journal of Population, 8(2), pp. 125–148.

Black

(2001) Environmental refugees: myth or reality? Working Paper No. 34, New Issues in Refugee Research, UNHCR.

Bongaarts

Frank

Lesthaeghe

(1984) The proximate determinants of fertility in sub-Saharan Africa, Population and Development Review, 10(3), pp. 511–537.

10.

Bridges

Oldeman

(1999) Global assessment of human-induced soil degradation, Arid Land Research and Management, 13(4), pp. 319–325.

11.

Brutzkus

(1975) Centralized versus decentralized pattern of urbanization in developing countries: an attempt to elucidate a guideline principle, Economic Development and Cultural Change, 23(4), pp. 633–652.

12.

Cairns

(2004) Sustainability ethics: world population growth and migration, Mankind Quarterly, 45(2), pp. 169–194.

13.

Carr

E. R.

(2005) Placing the environment in migration: environment, economy, and power in Ghana’s central region, Environment and Planning, 37(5), pp. 925–946.

14.

Chalmers

Hedges

Cooper

(2002) A brief history of research synthesis, Evaluation & the Health Professions, 25(1), pp. 12–37.

15.

Chen

Valente

Zlotnik

(1998) What do we know about recent trends in urbanization?, in: Bilsborrow

(Ed.) Migration, Urbanization, and Development: New Directions and Issues, pp. 59–88. Dordrecht: Kluwer Academic Publishers.

16.

Chen

Z. Y.

Zong

Y. Q.

Chen

Wang

(2008) Migration patterns of Neolithic settlements on the abandoned Yellow and Yangtze River deltas of China, Quaternary Research, 70(2), pp. 301–314.

17.

Corbett

(1988) Famine and household coping strategies, World Development, 16(9), pp. 1099–1112.

18.

Crane

Kinzig

(2005) Nature in the metropolis, Science, 308(5726), p. 1225.

19.

Cutler

(1986) The response to drought of Beja famine refugees in Sudan, Disasters, 10(3), pp. 181–188.

20.

DeMenocal

(2001) Cultural responses to climate change during the late Holocene, Science, 292(5517), p. 667.

21.

Doos

B. R.

(1994) Environmental degradation, global food production and risk for large-scale migrations, Ambio, 23(2), pp. 124–130.

22.

Erway Morinière

L. C.

(2009) Tracing the footprint of environmental migrants through 50 years of literature, in: Oliver-Smith

Shen

(Eds) Linking Environmental Change, Migration & Social Vulnerability, pp. 22–29. Source: Studies of the University: Research Counsel and Education, No. 12/2009. Institute for Environment and Human Security, United Nations University, Bonn.

23.

Ezra

Kiros

G.-E.

(2001) Rural out- migration in the drought prone areas of Ethipia: a multilevel analysis, International Migration Review, 35(3), pp. 749–771.

24.

Giosan

Filip

Constantinescu

(2009) Was the Black Sea catastrophically flooded in the early Holocene?, Quaternary Science Reviews, 28(1/2), pp. 1–6.

25.

Glass

(1976) Primary, secondary, and meta-analysis of research, Educational Researcher, 5(10), p. 3.

26.

Gonzalez-Samperiz

Utrilla

Mazo

. (2009) Patterns of human occupation during the early Holocene in the central Ebro basin (NE Spain) in response to the 8.2 ka climatic event, Quaternary Research, 71(2), pp. 121–132.

27.

Grimm

N. B.

Foster

Groffman

. (2008) The changing landscape: ecosystem responses to urbanization and pollution across climatic and societal gradients, Frontiers in Ecology and the Environment, 6(5), pp. 264–272.

28.

Haase

(2008) Urban ecology of shrinking cities: an unrecognized opportunity?, Nature and Culture, 3(1), pp. 1–8.

29.

International Federation of the Red Cross (2001) Climate migrants report.

30.

IPCC (Intergovernmental Panel on Climate Change) (2007) Climate change 2007: the physical science basis, summary for policy makers. Working Group 1, Fourth Assessment Report, IPCC, Geneva.

31.

Kabisch

Haase

(2006) Beyond growth: urban development in shrinking cities as a challenge for modeling approaches. International Environmental Modelling and Software Society ( http://www.iemss.org/iemss2006/papers/s8/S8_Kabisch.pdf ).

32.

Kasarda

J. D.

Cremshaw

E. M.

(1991) Third world urbanization: dimensions, theories and determinants, Annual Review of Sociology, 17, pp. 467–501.

33.

Kitchenham

(2004) Procedures for performing systematic reviews. Technical Report No. TR/SE0401, Keele University.

34.

Kitchenham

Pearl Brereton

Budgen

. (2009) Systematic literature reviews in software engineering: a systematic literature review, Information and Software Technology 51(1), pp. 7–15.

35.

Lawrence

Després

(2004) Futures of transdisciplinarity, Futures, 36(4), pp. 397–405.

36.

Lee

Farber

(1984) Fertility adaptation by rural–urban migrants in developing countries: the case of Korea, Population Studies: A Journal of Demography, 38(1), pp. 141–155.

37.

Lee

E. S.

(1966) A theory of migration, Demography, 3(1), pp. 47–57.

38.

Lein

(2000) Hazards and forced migration in Bangladesh, Norwegian Journal of Geography, 54(3), pp. 122–127.

39.

Lonergan

(1998) The role of environmental degradation in population displacement, Environmental Change and Security Project Report, 4, pp. 5–15.

40.

Main

H. A. C.

(1990) Dam projects and urbanization in Kano state, Nigeria, Singapore Journal of Tropical Geography, 11(2), pp. 87–99.

41.

Martine

(2009) Population dynamics and policies in the context of global climate change. Paper prepared for the Expert Group Meeting on Population Dynamics and Climate Change, UNFPA and IIED in collaboration with UN-HABITAT and the Population Division, UN/DESA, London, June.

42.

Max-Neef

M. A.

(2005) Foundations of transdisciplinarity, Ecological Economics, 53(1), pp. 5–16.

43.

McLeman

(2006) Migration out of 1930s rural eastern Oklahoma: insights for climate change research, Great Plains Quarterly, 26(1), pp. 27–40.

44.

McNamara

K. E.

(2007) Conceptualizing discourses on environmental refugees at the United Nations, Population and Environment, 29(1), pp. 12–24.

45.

Montgomery

(2008) The urban transformation of the developing world, Science, 319(5864), p. 761.

46.

Mougeot

(1992) Outmigration induced by environmental degradation. The Environment Department, World Bank, Washington, DC.

47.

Mulvihill

P. R.

Ali

S. H.

(2007) Disaster incubation, cumulative impacts and the urban/ ex-urban/rural dynamic, Environmental Impact Assessment Review, 27(4), pp. 343–358.

48.

Myers

(2002) Human well-being and the natural environment, Nature, 416(6883), pp. 788–789.

49.

Newland

(1980) City limits: emerging constraints on urban growth. Worldwatch Paper No. 38, Worldwatch Institute, Washington, DC.

50.

Osterling

J. P.

(1979) 1970 Peruvian disaster and the spontaneous relocation of some of its victims: Ancashino peasant migrants in Huayopampa, Mass Emergencies, 4(2), pp. 117–120.

51.

Pallagst

Aber

Audirac

. (2009) The future of shrinking cities: problems, patterns and strategies of urban transformation in a global context. Monograph No. 2009-01, Center for Global Metropolitan Studies, University of California (http://iurd.berkeley.edu/sites/default/files/pubs/MG-2009-01.pdf).

52.

Poulain

Perrin

(2009) Measuring international migration: a challenge for demographers, in: Surkyn

Deboosere

van Bavel

(Eds) Demographic Challenges for the 21st Century: A State of the Art in Demography, pp. 143–173. Brussels: VUBPRESS.

53.

Ravenstein

E. G.

(1885) The laws of migration, Journal of the Statistical Society of London, 48(2), pp. 167–235.

54.

Rees

W. E.

(1992) Ecological footprints and appropriated carrying capacity: what urban economics leaves out, Environment and Urbanization, 4(2), pp. 121–130.

55.

Rees

W. E.

(1996) Revisiting carrying capacity: area-based indicators of sustainability, Population and Environment, 17(3), pp. 195–215.

56.

Rees

W. E.

(1997) Is ‘sustainable city’ an oxymoron?, Local Environment, 2(3), pp. 303–310.

57.

Rees

W. E.

(2005) Ecological footprints and appropriated carrying capacity: what urban economics leaves out, Sustainability: Sustainability indicators, 4(2), p. 137.

58.

Renaud

(1981) National Urbanization Policy in Developing Countries. New York: Oxford University Press.

59.

Rieniets

(2009) Shrinking cities: causes and effects of urban population losses in the twentieth century, Nature and Culture, 4(3), pp. 231–254.

60.

Robinson

D. T.

Brown

D. G.

(2009) Evaluating the effects of land-use development policies on ex-urban forest cover: an integrated agent-based GIS approach, International Journal of Geographical Information Science, 23(9), pp. 1211–1232.

61.

Rosenfield

P. L.

(1992) The potential of transdisciplinary research for sustaining and extending linkages between the health and social science, Social Science & Medicine, 35(11), pp. 1343–1357.

62.

Rowell

E. J.

(1936) Drought refugee and labor migration to California in 1936, Monthly Labor Review, 43(6), pp. 1355–1363.

63.

Satterthwaite

(2009) The implications of population growth and urbanization for climate change, Environment and Urbanization, 21(2), pp. 545–567.

64.

Saxena

K. B.

(2008) Development, displacement, and resistance: the law and the policy on land acquisition, Social Change, 38(3), pp. 351–410.

65.

Scheper-Hughes

(1997) Demography without numbers, in: Kertzer

D. I.

Fricke

(Eds) Anthropological Demography: Toward a New Synthesis, pp. 201–222. Chicago, IL: University of Chicago Press.

66.

Taylor

P. S.

Vasey

(1936) Drought refugee and labor migration to California, June– December 1935, Monthly Labor Review, 42(2), pp. 312–318.

67.

Tibaijuka

(2010) Adapting to urban displacement, Forced Migration Review, 34(February), pp. 4–49.

68.

Timberlake

(1987) World-system theory and the study of comparative urbanization, in: Smith

Feagin

(Eds) The Capitalist City: Global Restructuring and Community Politics, pp. 37–65. London: Blackwell.

69.

Todaro

(1997) Urbanization, unemployment, and migration in Africa: theory and practice. Policy Research Division Working Paper No. 104 [published in D. Ghai (Ed.) (2010) Reviewing Social and Economic Progres in Africa: Essays in Memory of Philip Ndegwa, ch. 3. London: Macmillan].

70.

Turney

C. S. M.

Brown

(2007) Catastrophic early Holocene sea level rise, human migration and the Neolithic transition in Europe, Quaternary Science Reviews, 26(17/18), pp. 2036–2041.

71.

UNDESA (United Nations Department of Economic and Social Affairs) (1985) Urbanization prospects and database. UNDESA.

72.

United Nations (1982) Demopaedia: multilingual demographic dictionary, 2nd edn (http://en-ii.demopaedia.org/wiki/31).

73.

United Nations (2008) World urbanization prospects: the 2007 revision. Population Division, Department of Economic and Social Affairs.

74.

UNOCHA (United Nations Office for the Co-ordination of Humanitarian Affairs) and IDMC (Internal Displacement Monitoring Centre) (2009) Monitoring disaster displacement in the context of climate change. IDMC, September.

75.

Waal

A. de

el Amin

(1986) Survival in northern Darfur 1985–1986. Save the Children Fund, Nyala, Sudan.

76.

Watts

M. J.

Bohle

H. G.

(1993) The space of vulnerability: the causal structure of hunger and famine, Progress in Human Geography, 17, pp. 43–67.

77.

Wood

W. B.

(2001) Ecomigration: linkages between environmental change and migration, in: Zolberg

A. R.

Benda

(Eds) Global Migrants, Global Refugees: Problems and Solutions, pp. 42–61. New York: Berghahn Books.

78.

Zetter

Deikun

(2010) Meeting humanitarian challenges in urban areas, Forced Migration Review, 34(February), pp. 5–8.